MindPortal

Inactive

Thought-to-Language AI Models

Winter 2021Founded 201910 peopleLondon, England, United Kingdom

www.mindportal.com ↗LinkedIn ↗See on the Idea Map Consumer momentum

About

MindPortal is a pioneering AI research company advancing human-AI interaction by enabling people to communicate with computers and AI assistants purely through thought. As an AI company, we focus exclusively on model development rather than hardware. Our models are designed to be general-purpose, adaptable across a wide range of applications. Our flagship development, MindSpeech, is the first supervised AI model to translate imagined speech into text, enabling direct thought-to-language communication. MindSpeech has earned endorsements from leading experts, including Neuralink's former technical strategy head and DARPA’s N3 program lead scientist. Our thought-to-language AI models have the potential for transformative applications in fields such as healthcare, education, and VR/AR and our work has been featured in major publications such as Forbes, The Times, The Telegraph, and Business Insider. MindPortal has over 11 patents filed in the field and is backed by prominent investors such as Y Combinator.

From their website

as of Jun 7, 2026www.mindportal.com ↗

API/Infra

MindPortal claims to have built telepathic AI models that interpret human thoughts to enable thought-based interaction with AI, including a thought-to-text model (MindSpeech), a thought-based LLM interface (MindGPT), and a thought-based GUI selector (MindClick). They position these as ready-to-scale research models backed by Y Combinator.

MindSpeech decodes free-form thoughts into text without templates or memorization, generalizes to new users, and works across hundreds of topics. MindGPT enables communication with AI using only thoughts by routing pre-memorised sentences to an LLM via thought alone, with direct integration to an AI like ChatGPT. MindClick allows thought-based GUI selections, designed to work with VR/AR hardware and computer screens, providing rapid response times. The models are presented as ready-to-scale research prototypes with a focus on hands-free, silent, privacy-centric interfaces.

Who it’s for: Potential developers, AR/VR hardware makers, accessibility-focused organizations, and researchers aiming to deploy or integrate thought-based interfaces at scale.

Features